Ž .Brain Research 792 1998 193–206

Research report

Substance P attenuates and DAMGO potentiates amygdala glutamatergicneurotransmission within the ventral pallidum

Igor Mitrovic 1, T. Celeste Napier )

Department of Pharmacology and Experimental Therapeutics, Loyola UniÕersity Chicago, Stritch School of Medicine, Maywood, IL 60153, USA

Accepted 27 January 1998

Abstract

Ž . Ž . Ž .The amygdala AMG , nucleus accumbens NA and ventral pallidum VP influence goal-oriented behaviors. However, the nature ofthe interactions among these regions has not been well characterized. Anatomical studies indicate that excitatory amino acids are

Ž .contained in VP inputs from the AMG, and the NA is a primary source of VP substance P SP and opioids. The present study wasdesigned to functionally characterize the NA and AMG projections to the VP, and to assess if opioids and SP can modulateAMG-mediated excitatory neurotransmission within the VP. To do so, extracellularly recorded electrophysiological responses of singleVP neurons to electrical activation of VP afferents were monitored during microiontophoretic application of treatment ligands in chloralhydrate-anesthetized rats. The anatomically described glutamatergic inputs from the AMG, and SP inputs from the NA, werepharmacologically verified. It also was determined that even though iontophoretically applied SP increased the spontaneous activity of VP

Ž .neurons, at ejection current levels that were below those necessary to produce this effect termed sub-threshold , the tachykinin attenuatedAMG stimulation-evoked glutamatergic neurotransmission. SP failed to modulate the excitations induced by iontophoretically appliedglutamate suggesting that SP modulation of AMG-evoked excitations were mediated via a decrease in the pre-synaptic release ofglutamate. Like SP, the effects of sub-threshold ejection currents of m opioid agonist DAMGO on AMG-evoked responses were notpredicted by the opioid’s effects on spontaneous VP neuronal activity; DAMGO inhibited spontaneous firing but potentiated AMG-evokedglutamatergic neurotransmission. The opioid also potentiated effects of exogenous glutamate implying an interaction at a post-synapticsite. These results indicate that tachykinin and opioid neuropeptides contained in NA projection neurons can differentially modulate AMGglutamatergic inputs to the VP. q 1998 Elsevier Science B.V.

Keywords: Ventral Pallidum; Amygdala; Nucleus accumbens; Opioid; DAMGO; Tachykinin; Substance P; Excitatory amino acid; Glutamate; Neuromodu-lation

1. Introduction

Anatomical studies using double-labeling immunohisto-Ž .chemical techniques demonstrate that substance P SP and

enkephalins are present in ventral striato-pallidal projec-w xtion neurons 15,17,45,51 . Lesions of the nucleus accum-

Ž .bens NA almost completely deplete SP and enkephalinŽ .immunoreactivity in the ventral pallidum VP suggesting

Ž .that the NA may be the main if not the only source of SP

) Corresponding author. Department of Pharmacology and Experimen-tal Therapeutics, Loyola University Chicago, Stritch School of Medicine,2160 S. First Ave., Maywood IL 60153, USA. Fax: q1-708-216-6596;E-mail: cnapier@luc.edu

1 Current address: Departments of Anatomy and Neurology, KeckCenter for Integrative Neuroscience, University of California at SanFrancisco, 513 Parnassus Ave., Box 0452, San Francisco, CA 94143,USA.

w xand opioid inputs to the VP 17,62 . The NK tachykinin1

receptor, and m, d and k opioid receptors are foundw xwithin the VP 24,27,30,38,50 . Activation of these recep-

tors via microiontophoretic administration of tachykininand opioid agonists has demonstrated their importance in

w xthe regulation of VP neuronal activity 8,34,35,43,45 .Among the opioid receptor subtypes, m activation appears

w xto be particularly important for VP function 3,19,34,42 .Additionally, microiontophoresis of the opioid antagonistnaloxone attenuates VP neuronal responses evoked byelectrical stimulation of NA, thus verifying the involve-ment of opioid neuropeptides in NA-to-VP transmissionw x8 . A functional verification for the presence of SP in thisprojection has not been conducted; thus this was one of theobjectives of the present study.

Anatomical studies employing a variety of track-tracingtechniques have demonstrated a projection from the AMG

0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII S0006-8993 98 00130-9

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206194

to basal forebrain regions, including the VPw x13,20,23,33,63 , and it is likely that the excitatory amino

w xacid glutamate is utilized by this projection 13 . Microion-tophoretically applied glutamate increases neuronal activ-

w xity in over 90% of the VP neurons tested 41 , andŽ .electrical stimulation of the amygdala AMG orthodromi-

Ž .cally evokes short latency F12 ms excitatory responsesw xin the VP 32,57,61 . These observations suggest that the

AMG-evoked responses are most likely monosynaptic andmediated by glutamate. An objective of the present studywas to verify this conclusion.

The AMG, NA, and VP regulate several behavioralfunctions including motor, reward, cognition, and affective

w xstates 1,25,44,47 . Interactions among these forebrain re-gions may be critical for the generation and execution ofcomplex behavioral patterns such as goal-directed behav-

w xiors 36 . NA and AMG efferents converge in the VP. Apossible mode of interaction between these two VP affer-ents may involve the ability of SP and opioids to modulate

w xexcitatory amino acid-mediated transmission 46,53 . Themain objective of the present study was to examine thesepossibilities at the level of the single VP neuron.

To meet the study’s objectives, we conducted electro-physiological experiments in anesthetized rats with extra-cellular recordings from single VP neurons during electri-cal stimulation of the AMG and microiontophoretic appli-cation of ligands. This experimental approach enabled us:.1 to stimulate specific VP afferents and study the effects

of endogenously released neurotransmitters on single VP.neurons in a virtually intact system; 2 to apply discrete

amounts of selective antagonists within the local environ-ment of the recorded VP neuron in order to identify theneurotransmitter mediating the stimulation-evoked re-sponses by the ability of its antagonist to block these

.responses; 3 to compare the effects of exogenously ap-Ž .plied via microiontophoresis and endogenously released

Ž .via electrical stimulation neurotransmitters on single VP.neurons; and 4 because of a relatively quick onset and

offset of effects elicited by iontophoretically applied lig-ands, we could repeatedly test exogenous compounds totitrate the amount of a ligand used. This later aspect wascritical to study modulation for it allowed us to investigatethe ability of potential modulators to alter neuronal firingelicited by electrical stimulation of the AMG while exert-

Ž .ing a minimal effect or no effect on the spontaneous VPneuronal firing. A portion of these data was published inabstract form in Society for Neuroscience Abstracts21:1600, 1995.

2. Materials and methods

2.1. Animal surgery

Animals were handled in accordance with proceduresrecommended in Guide for the Care and Use of Labora-

tory Animals published by Institute of Laboratory AnimalResources, Commission on Life Sciences, and the NationalResearch Council.

Ž .Male Sprague–Dawley rats Harlan, Indianapolis, INweighing 280–350 g were anesthetized with chloral hy-

Ž .drate 400 mgrkg, i.p.; Sigma, St. Louis, MO . A lateraltail vein was cannulated for intravenous administration ofthe anesthetic to maintain surgical level of anesthesia. The

Žrats were mounted into a stereotaxic apparatus David.Kopf Instruments, Tujunga, CA with the nose piece set at

y3.3 mm. The skull was exposed and holes were drilledthrough the skull overlying the VP and either the AMG orthe NA. Stereotaxic coordinates used for recording from

Ž .the VP were 0.5 mm posterior to bregma P , 2.5 mmŽ . Ž .lateral to the midline L , and 7.3–8.3 mm below dura V .

A stimulating electrode was placed at 1.2 mm anterior toŽ .bregma A , 1.5 mm L and 6.8 V for NA stimulation, and

2.8 mm P, 4.7 mm L and 7.6 mm V to activate the AMG.Throughout the experiment, the rat’s body temperature wasmaintained at 35–378C with a thermostatically controlled

Ž .heating pad Fintronics, Orange, CT .

2.2. Micropipettes and drugs

w xA multibarrel pipette-microelectrode assembly 11 , wasused for the microiontophoretic application of ligands andextracellular isolation and monitoring of the individualaction potentials. This assembly provides stable recordingsof VP neuronal activity, and consistent responses to mi-

w xcroiontophoretically applied ligands 34,41 . To constructthis pipette-microelectrode, single and five-barrel glass

Žpipettes glass tubing purchased from A-M Systems, Ev-. Žerett, WA were heat-pulled vertical puller, Narishige

.PE-2, Tokyo, Japan and broken back to 2 to 3 mm and 9to 11 mm OD, respectively. The single pipette served asthe recording microelectrode, and was glued in parallel tothe five-barrel iontophoretic pipette so that the microelec-trode tip extended 10–20 mm beyond the iontophoreticpipette. The recording microelectrode and the current-bal-ancing barrel of the iontophoretic pipette were filled with2% pontamine sky blue dissolved in a 0.5 M solution ofsodium-acetate. The remaining four barrels of the ion-tophoretic pipette contained different combinations of the

w 2 4 5 xfollowing: DAMGO D-Ala ,n-Me-Phe ,Gly-ol -En-Ž .kephalin Bachem, Torrance, CA; 10 mM in H O , a2

w xhighly selective m opioid agonist 10 ; substance PŽ .Bachem; 1 mM in H O , a NK receptor-preferring2 1

w x Ž . Žtachykinin 52 ; CP96345 2 S,3S -cis-2- diphenyl-. ŽŽ . . w xmethyl -N- 2-methoxyphenyl methyl -1-azabicyclo- 2.2.2

Žoctan-3-amine dihydrochloride Pfizer, Groton, CT; 10.mM in H O , a selective NK tachykinin receptor antago-2 1

w x Ž .nist 52,56 , glutamate Sigma, 200 mM in H O , and2Žkynurenic acid KYN; 50 mM in H O, pH adjusted to2

. w x8.0 , an excitatory amino acid antagonist 9 . The in vitroimpedances of the recording electrodes were between 3and 6.5 MV when measured at 165 Hz with a microelec-

Ž .trode tester Winston Electronics, San Francisco, CA .

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206 195

Similarly measured impedances of the iontophoreticpipettes ranged from 20–85 MV, and those of the currentbalancing barrel ranged from 8–20 MV.

2.3. Microiontophoresis

The pipette assembly was lowered in the brain with aŽ .hydraulic microdrive Trent-Wells, South Gate, CA .

Spontaneous VP neuronal activity was sampled, and moni-Žtored with a storage oscilloscope Kikusui Electronics,

. ŽJapan and an audio monitor Grass Instruments, Quincy,.MA . Individual action potentials were amplified, filtered

Ž .200 Hz and 2 kHz , and isolated from the backgroundŽ .with an amplifierrvoltage discriminator Fintronics . The

output was transferred to an IBM-compatible PC, which,using custom software, generated on-line histograms andquantified the data. Using a six-channel current generator

Ž .and programmer Fintronics , drugs were microion-tophoretically ejected using currents that ranged from 1–120 nA, and retained with 10 nA. Glutamate was ejectedwith an anionic current and retained with a cationic cur-rent; the opposite was used for ejection and retention of allthe other drugs. At the beginning of each electrode pene-

Ž .tration while dorsal to VP , ejection currents were appliedto each of the drug-containing barrels for 30 min toconcentrate the drugs at the pipette tip. Subsequently, aspontaneously firing neuron was isolated and monitored,and a stable recording rate was obtained from whichbaseline firing frequency and firing pattern were deter-mined. For those cells with consistent interspike intervals,a stable recording was achieved when firing variation wasless than 15% for at least 3 min. Because some VP cells

Ž .fire in a periodic i.e., cyclic pattern, stable firing wasconsidered to have occurred with these neurons only if the

Žmean firing for each cycle of at least three cycles or 3 min.of recording, depending upon which was longer , did not

vary more than 15%.Changes in neuronal activity were attributed to the

iontophoretically applied drug if the ejection currents of nomore than 40 nA uniformly altered the firing rate by atleast 20% from the pretreatment rates during at least two

Ždrug applications each with a duration of at least two.min . After meeting this criterion, drugs were microion-

tophoresed from the pipette using 5–20 s ejectionr30 sretention epochs for glutamate, and a minimum of 1 minejection periods for the remaining drugs. Drug-inducedeffects were evaluated by comparing neuronal firing, as

Ž .averaged across at least three rate-sampling periods binsto equal or exceed a total of 15 s, at the end of the drugapplication period, with firing rate that immediately pre-

Žceded the ejection of the drug using the same protocol to.determine rate average . This assessment was aided by a

visual inspection of the real time-rate histograms to deter-mine consistency in firing pattern during treatment applica-tions and to verify that post-treatment recovery of baselinerate was obtained. With those neurons where periodicfiring pattern occurred, care was taken to administer all

treatments during the same phase of the cycle. This mi-croiontophoresis protocol has been demonstrated to consis-tently provide reliable measures of the drug-induced ef-

w xfects in VP neuronal activity 34,35 .

2.4. Stimulation paradigms and study of modulation

For electrical stimulation of the AMG and the NA,Žstainless steel concentric bipolar electrodes NEX-100; 0.5

mm OD, 0.2 mm ID, with 0.5 mm exposed tips separated.by 0.5 mm; David Kopf Instruments were used to deliver

current generated by a Grass S88 stimulator coupled to aŽ .Grass stimulus isolation unit SIU 5 , and a Grass constant

Ž .current unit CCU 1 . Single, 0.1 ms monophasic pulses ofvarious current strengths were applied at 1 Hz to stimulatethe AMG and the NA. To assure consistency amongrecorded neurons, the effects of agonists and antagonistswere evaluated using stimulating currents that produced50–75% of the maximum evoked response for the shortlatency components.

A peristimulus-time histogram was generated for eachsample of 128, 1 Hz stimulation epochs, using a bin widthof 2 ms. To quantify stimulation evoked responses andtheir modulation by neuropeptides, the following assess-ments of the peristimulus time histograms were conducted:

Ž .The mean number of action potentials or counts per binŽ .occurring during 40 bins 80 ms immediately preceding

the stimulation was considered as the baseline and theaverage counts per bin for this time period was used to

Ž .determine the prestimulus mean. The 50 bins 100 msimmediately following the stimulation were analyzed forthe occurrence of the evoked responses. The onset and

Žoffset of an evoked response component inhibition or.excitation were characterized for the 128 epochs during

Ž .which no drugs were applied control . The onset of theevoked response was defined as the first of three consecu-tive poststimulus bins that differed from the prestimulusmean by more than 1.26 S.D.; offset was defined by thefirst of three consecutive bins that no longer met this

w xcriterion. As described by Chrobak and Napier 8 , thisapproach sets p at -0.001. Counts occurring within acomponent of the evoked response during control werecompared to the number of counts occurring within thesame component during iontophoretic application of atreatment drug. The microiontophoresed drugs and their

.corresponding stimulation site were: 1 DAMGO applied.during stimulation of the AMG; 2 SP applied during

.stimulation of AMG; 3 KYN applied during stimulation.of AMG; and, 4 CP96345 applied during stimulation of

NA. In some VP neurons, microiontophoresis of some ofthese drugs were sufficient to slightly alter the interstimu-

Ž .lus baseline as well as spontaneous firing . To assure thateven small treatment-induced alterations in interstimulusbaseline would not produce false positives for comparisonsof evoked responses between the treatment and controlconditions, we used a previously established method for

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206196

data analysis that eliminates the influence of this con-founding variable. This method predicts the number ofaction potentials that would occur in each evoked responsecomponent during control if the control interstimulus base-line was the same as the one obtained during treatmentw x8,32,40 . This ‘expected number’ was determined by mul-tiplying the number of counts occurring within an evoked

Ž .component of the control sample no drugs iontophoresedby the ratio of the treatment prestimulus mean to thecontrol prestimulus mean. Thus, by taking into accountchanges in interstimulus baseline and the number of countsoccurring within an evoked component of the controlsample, determining the expected number allows the ‘iso-lation’ of the effects of treatment on the evoked responseregardless of its effects on the interstimulus baseline.Predictably, when calculated for neurons whose interstimu-lus baseline was not altered by the treatment, the expectednumber did not differ from the actual number of countsoccurring within an evoked component during the controlŽ .examples are provided in the legends for Figs. 5 and 6 .Thus, for statistical analysis, the expected number wascalculated for all VP neurons tested. To determine if drugtreatment significantly altered evoked responses, the ex-pected numbers for control periods were compared withthe number of counts that actually occurred in the sameevoked component during treatment using a paired t-test.

Some neurons responding to the stimulation of theAMG with an excitatory response were evaluated to deter-mine if glutamate mediated this response. To do this, theAMG was stimulated using currents that produced 50–75%of the maximum evoked response while KYN was ion-tophoresed at an ejection current level demonstrated toantagonize neuronal responses to iontophoretically appliedglutamate. Because antagonists have no efficacy, antago-

Žnism of a short latency excitatory response which is most.likely monosynaptic evoked by AMG stimulation would

indicate that this evoked component is elicited via anactivation of the direct glutamatergic projection from theAMG. The contribution of the NA to SP innervation of theVP was evaluated in neurons excited by both electricalstimulation of the NA and iontophoretically applied SP.This was accomplished by evaluating the ability ofCP96345 to antagonize the excitatory evoked responseselicited by the stimulation of the NA using the protocolemployed to demonstrate the glutamatergic nature of theAMG-evoked responses.

To investigate the capability of iontophoretically ap-plied DAMGO and SP to modulate neuronal responseselicited by electrical stimulation of the AMG, the follow-

.ing protocol was used: 1 a spontaneously active VPneuron was isolated and a stable baseline firing rate was

. .obtained; 2 the AMG was electrically stimulated; 3 neu-rons responding to AMG stimulation were tested for sensi-

.tivity to iontophoretically applied agonists; and, 4 neuronssensitive to an agonist were tested for agonist modulationof the AMG-evoked response. Modulation was assessed

.using the following parameters: 1 To determine the ejec-tion current level for the agonist, starting from an ion-tophoretic current that produced a robust neuronal re-sponse, the ejection current was decreased until the neu-ronal activity was altered less than 20% from baselinefiring rate. This ‘sub-threshold’ ejection current was used

.for subsequent evaluations. 2 The stimulation currentlevel for the AMG was selected by testing a range ofcurrent magnitudes, and the stimulating current that pro-duced 50–75% of the maximally-evoked response wasused for subsequent evaluations. This approach assuredthat both an augmentation or a suppression of the AMGstimulation-evoked response by SP andror DAMGO couldbe observed. When possible, 4–12 min after termination ofthe agonist ejection current, electrical stimulation of AMGwas repeated to test for the recovery of the AMG-evokedresponse.

2.5. Histological eÕaluations and statistics

At the end of each experiment, an anionic current waspassed through the recording microelectrode to depositpontamine sky blue and mark the recording site. Theanimal then was deeply anesthetized with chloral hydrate,and perfused with 0.9% sodium-chloride. The brain wasremoved from the skull and stored in a 4% formalin-20%sucrose solution for fixation. Fixed brains were cut with a

Ž .cryostat—microtome Hacker Instruments, Fairfield, NJ .Sections containing blue dye deposits, and microelectrodeand stimulating electrode tracks, were mounted on speci-men glass slides and stained with neutral red. After agree-ment by two persons on the locations of the blue dots, andthe stimulating electrode tips, the placements were marked

w xon stereotaxic maps of the rat brain 49 . The blue dotserved as a reference point for stereotaxic localization ofthe remaining recording sites.

Table 1Distribution of VP responses according to the evoked response onset latency and direction

Ž . Ž .Nucleus stimulated VP neurons tested Short latency response F12 ms Long latency response )12 ms Not evoked

Excitation Inhibition Excitation Inhibition

AMG 105 40 17 29 18 17NA 24 6 2 10 6 5

Sixteen neurons evoked by AMG stimulation, and five neurons evoked by NA stimulation demonstrated complex responses consisting of more than oneevoked component. Therefore the numbers do not add up to 105 and 24.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206 197

Ž .Fig. 1. Peristimulus histograms 128 epochs each illustrating thecurrent-dependency of the complex responses evoked in VP neurons byelectrical stimulation of the AMG. Horizontal bars flanked with arrowsindicate statistically determined onset and offset of evoked componentsŽ .see Section 2 . Numbers to the left of the stimulus artifact indicate themagnitude of the AMG stimulating current. As illustrated by this exam-ple, in some neurons, increasing stimulation current magnitude increased

Ž .the complexity of the evoked response. Note bottom histogram how aŽ .long latency )12 ms inhibitory response separates what appeared to beŽ . Ž .short latency F12 ms excitation of long duration 52 ms into a short

Žlatency excitation of short duration typical excitatory response to AMG.stimulation , and a long latency excitation.

To describe the relationship between the location of theVP neurons and their sensitivity to stimulation of AMG,NA and iontophoretically applied drugs, the VP was di-

wvided into five subregions as follows the coordinatesclosely correspond to those illustrated by plates within

w x .Paxinos and Watson 49 : 1 the rostral VP immediately infront of the anterior commissure, crossing at approximately

.0.05 mm anterior to bregma; 2 the subcommissural VP atthe level of the crossing of anterior commissure, 0.26 mm

.caudal to bregma; 3 the subcommissural VP at 0.40 mm.caudal to bregma; 4 the rostral substantia innominata and

.the VP at 0.80 mm caudal to bregma; 5 the rostralsubstantia innominata and the VP at 0.92 mm caudal tobregma. The rostral three subregions were divided furtherin ventromedial and dorsolateral territories according to

differences in histochemical markers and circuitry profilesw x15,65,66 .

For analysis of drug effects on the VP neuronal activity,pretreatment firing rate was standardized to 0%. Drugeffects are reported as a percentage change from thepretreatment rate. Student’s t-test, paired t-test and x 2-testwere used for the data comparisons with a pF0.05required for significance. All parametric data are presentedas mean"S.E.M.

Ž .Fig. 2. Location of the stimulation sites within the NA and the AMG AŽ .and the recording sites within the VP B . Numbers in the upper right

Ž .corner of each plate represent distances from bregma in mm . A. Thestars indicate the site of a stimulating electrode tip. Abbreviations: BLA,basolateral amygdala; BLP, basolateral posterior amygdala; CMA, centro-medial amygdala; NAC, nucleus accumbens core; NAS, nucleus accum-bens shell. B. Stereotaxic maps illustrating the recording sites for VPneurons recorded during AMG stimulation. Neurons represented withopen diamonds were not sensitive to AMG stimulation; filled diamondsrepresent neurons sensitive to AMG stimulation but not tested with eitherSP or DAMGO. All subsequent neurons were sensitive to AMG stimula-tion and tested for sensitivity to SP and DAMGO. Filled circles representneurons sensitive to both SP and DAMGO; neurons not sensitive to eitherSP or DAMGO are represented with open circles; filled squares representneurons sensitive to SP only; open squares represent neurons not sensitiveto SP; neurons sensitive only to DAMGO are represented with filledtriangles; open triangles represent neurons not sensitive to DAMGO.Abbreviations: ac, anterior commissure; Fstr, striatal fundus; GP, globuspallidus; HDB, horizontal limb of the diagonal band; MCPO, magnocellu-lar preoptic nucleus; SI, substantia innominata; VP, ventral pallidum.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206198

3. Results

One hundred thirty-one spontaneously active VP neu-rons with a firing rate of 10.3"0.7 Hz were investigatedin this study. Sixty-one percent of the VP neurons demon-strated biphasic action potentials and 39% were triphasic.Initially positive and negative deflections of the actionpotentials were encountered in almost equal proportions.The amplitude of the recorded action potentials was 418"

19 mV, and the duration was 1.8"0.1 ms.

3.1. Responses of VP neurons to stimulation of the AMGand the glutamatergic nature of the eÕoked excitations

Activity of 105 VP neurons was recorded during electri-cal stimulation of the AMG. Eighty-four percent responded

Ž .to the stimulation Table 1 , and both inhibitory andexcitatory responses were observed. One of the neuronsthat demonstrated an evoked response to AMG stimulationwas possibly activated antidromically because it was able

Ž .to follow high frequency 200 Hz pulse-pair stimulationand the evoked response occurred with constant onset

w xlatency 29 . This neuron was not studied further. Theremaining responses evoked by stimulation of the AMGdid not exhibit these characteristics, and thus were consid-ered to be orthodromically activated. VP neuronal re-sponses were dependent upon the magnitude of the stimu-lation current; i.e., higher stimulating currents often in-creased the duration andror the number of counts in an

Ž .evoked response component Fig. 1 . In some neurons,increasing the stimulating current also increased the com-

Ž .plexity of the long latency response pattern Fig. 1 .Excitatory responses evoked by AMG stimulation

demonstrated a skewed unimodal distribution of onsetlatencies. Fifty-six percent of all excitatory responses wereclustered between 4 and 12 ms. A peak in the distributionhistogram occurred at 12 ms, which was followed by asharp decline so that onset latencies for the remainingexcitatory responses were uniformly distributed between14 and 80 ms. Inhibitory responses exhibited a bimodaldistribution, with one peak occurring at 2–6 ms, and

Ž .Fig. 3. The effects of glutamate GLU on spontaneous neuronal activityŽ .in the VP and the ability of kynurenic acid KYN to antagonize

GLU-mediated excitations as well as the short latency excitations evokedŽby stimulating the AMG data represented in A and B were collected

.from the same neuron . A. A real time rate histogram demonstratingKYN antagonism of firing rate increases mediated by GLU. The dashes

Ž .indicate GLU application in 20 s pulses . Note the very short onset andoffset of the GLU-evoked responses. The horizontal bar flanked witharrows indicates onset and offset of ejection current used to apply KYN.The numbers above the lines are the magnitude of the iontophoretic

Ž .current. B. Peristimulus time histograms 128 epochs each illustratingthe ability of KYN to suppress excitation evoked by AMG stimulation.This result demonstrates that the response was mediated by glutamate.The number to the left of the stimulus artifact provides the stimulationcurrent magnitude applied to the AMG. The boxes indicate the onset,duration and peak countsrbin of excitation as defined during the control

Ž .condition i.e., before KYN iontophoresis; CTRL . KYN 120 nA indi-cates ejection current magnitude of kynurenic acid applied during AMGstimulation. In this neuron, 63 action potentials occurred in the evokedcomponent during CTRL. Thirty four action potentials occurred in thiscomponent during iontophoresis of KYN, and the expected count wascalculated to be 56.6 After recovery from KYN, the component contained64 action potentials. Although KYN slightly decreased number of action

Žpotentials counted in prestimulus baseline which is reflected in anexpected count of 56.6 versus 63 action potentials that actually occurredin the short latency excitation in CTRL, the much stronger attenuation of

Ž .the short latency excitation during KYN 34 counts demonstrates theability of KYN to antagonize the AMG-evoked excitations. C. A bargraph summarizing the suppression of the short latency responses toAMG stimulation by iontophoretically applied KYN. The open barŽ .CTRL represents the total number of counts in short latency excitationsevoked in VP neurons by AMG stimulating currents that produced

Ž .approximately 50% of the maximal response 0.4–1.4 mA . The stippledbar represents calculated expected number of counts for the short latency

Žresponses obtained during iontophoresis of KYN at 120 nA. ) p-0.05;paired t- test comparing short latency excitations during KYN iontophore-

w x .sis with expected number of counts defined in Section 2 for 7 neurons.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206 199

another at 16–21 ms. Based on these profiles, VP re-Ž .sponses were categorized as having a short F12 ms or

Ž . Ž .long )12 ms onset latency Table 1 . Short and longlatency excitatory responses exhibited durations of 16"3ms and 27"4 ms, respectively. Durations of inhibitoryresponses were 49"8 ms for the short latency componentand 39"7 ms for the long latency component. No correla-tion was found between different AMG stimulation-evokedresponses and any of the other extracellularly measuredelectrophysiological characteristics.

Most of the stimulation sites were within the basolateralŽ .nucleus of the AMG Fig. 2A . No trends in the VP

neuronal response profile could be attributed to the ros-tralrcaudal placement of the stimulating electrode withinthe AMG. Analysis of the recording electrode placementsrevealed that the entire ventro-dorsal and rostro-caudal

Žextent of the VP excluding the most caudal sublenticular. Ž .regions was sampled in this study Fig. 2B . Neither

Žcategory of the VP neuronal responses i.e., excitatory or.inhibitory and short or long onset latency was encoun-

tered with a significantly higher frequency either along therostral–caudal or the ventromedial–dorsolateral axis of theVP.

Close to 40% of the responses evoked by the AMGstimulation in the present study were excitatory with anonset latency which is suggestive of a monosynaptic inputŽ .F12 ms; see Table 1 . Seven VP neurons demonstratingshort latency excitations were tested with iontophoreticallyapplied glutamate and all seven were excited by the trans-

Ž .mitter e.g., Fig. 3A . KYN almost completely blockedŽeffects of iontophoretically applied glutamate ejection cur-

.rents ranging from 10–40 nA in each of these sevenŽ .neurons e.g., Fig. 3A . KYN also antagonized the AMG-

evoked short latency excitations at the same ejection cur-rent level that antagonized the excitatory effects of ion-

Žtophoretically applied glutamate in each of these cells Fig..3B and C .

3.2. Neuronal responses to stimulation of the NA and SPinÕolÕement in the eÕoked excitations

The involvement of opioid peptides in the NA-to-VPw xtransmission has been demonstrated previously 8 . The

present experiment concentrated on the contribution of SPto VP responses evoked by NA stimulation. Seventy onepercent of the 24 VP neurons tested were evoked by

Ž .stimulating the NA Table 1 . The responses demonstrateda variable onset latency and did not follow stimulation ofhigh frequencies indicating that they most likely were

w xmediated orthodromically 29 . Twenty-one percent of theevoked neurons exhibited a complex response pattern thatusually included both excitatory and inhibitory responses.Analysis of the stimulating electrode placements demon-strated that the core of the NA was stimulated almost

Ž .exclusively Fig. 2A . However, the relatively small num-ber of VP neurons tested with NA activation prevented an

analysis of correlation between different response cate-gories and location of the recorded neurons within differ-ent VP sub-territories. The distribution histogram for onsetlatencies for the excitatory responses was bimodal withone peak occurring between 4–12 ms, and another onebetween 28–40 ms. Onset latencies among eight neurons

Ž .responding with an inhibitory evoked component Table 1ranged from 4 to 96 ms without a peak in the distributionhistogram. Based on the latencies of the excitatory re-sponses in the present experiment, as well as in our

w xprevious study 8 , neuronal responses to NA stimulationŽ .were categorized as short latency F12 ms or long

Ž .latency )12 ms responses. Short latency excitatory re-sponses exhibited durations of 11"3 ms, and durations of

Fig. 4. Demonstration of the ability of the NK receptor antagonist1

CP96345 to antagonize the rate increases induced by SP iontophoresisŽand those evoked by NA stimulation the same cell is represented in A

.and B . A. A real time rate histogram illustrating antagonism of SP-in-duced responses by CP96345. The horizontal bars flanked with arrows

Ž .indicate onset and offset of the ejection currents. The dashed line stimindicates the time period during which the NA was stimulated. Thenumbers above the bars indicate magnitude of the iontophoretic ejection

Žused to apply the drugs. B. Peristimulus time histograms 128 epochs.each illustrating the ability of CP96345 to suppress VP excitatory

responses to stimulating the NA. The number to the left of the stimulusartifact illustrates the NA stimulation current. The boxes indicate onset,duration and the peak countrbin of excitation as defined during the

Ž .control condition i.e., before the application of CP96345; CTRL . In thisneuron, 60 action potentials occurred in the evoked component duringCTRL. The expected number was calculated to be 39.5, and the actualcount for the evoked component during CP96345 iontophoresis was 23.The NK antagonist reduced number of action potentials in prestimulus1

Žbaseline reflected in 60 action potentials occurring during CTRL versus.an expected number of 39.5 however, CP96345 had much stronger effect

Žon the evoked response illustrated by the 23 action potentials in the.evoked component during the iontophoresis of the drug indicating that

CP96345 was capable of attenuating NA-evoked excitations of VP neu-rons.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206200

the long latency excitatory responses were 46"19 ms.Durations for the short and the long latency inhibitoryresponses were 24"12 ms, and 121"34 ms, respec-tively. Extracellularly recorded electrophysiological neu-ronal characteristics did not differentiate among VP neu-rons exhibiting these various responses to NA stimulation.

Ten VP neurons demonstrating NA stimulation-evokedŽ .excitations onset latencies ranging from 4–90 ms were

tested for sensitivity to iontophoretically applied SP. Six ofthese were insensitive to the tachykinin, and in four cells,SP produced a slowly incrementing increase in firing. Thetachykinin NK receptor antagonist CP96345 attenuated1

Ž .this increase in each cell e.g., Fig. 4A . Also, NA activa-tion evoked only long latency excitations in the fourSP-sensitive cells, which concurs with the notion that SPmay mediate this excitatory component. To test this hy-pothesis, CP96345 was iontophoresis while the NA wasstimulated. The CP96345 ejection current that produced

Žantagonism of iontophoretically applied SP was used 25–.40 nA . Consistent with its ability to block SP, CP96345

abolished the long latency excitations in all four neuronsŽ .ts5.736; p-0.02; see Fig. 4B , thus verifying SP’sinvolvement in this evoked component. One of the fourneurons exhibited both short and long latency excitatorycomponents. In this cell, CP96345 decreased the longlatency excitatory component by 50%, but the short la-tency excitatory component was decreased by only 13%suggesting that a transmitter other than SP may mediatethe short onset latency response to NA stimulation.

3.3. The effects of SP and DAMGO on short latencyexcitations elicited by stimulation of the AMG

To assess the possibility that NA neuropeptides modu-late AMG transmission within the VP, 61 spontaneouslyfiring VP neurons that responded to AMG stimulationwere tested for responsivity to SP and DAMGO. Of theseneurons, 42 were tested for sensitivity to either SP orDAMGO, and 19 were tested with both peptides. Neuronalresponses to iontophoretically applied SP or DAMGOwere typically slow in onset and persisted after terminationof the iontophoretic current. In 37% of the neurons testedwith both peptides, spontaneous firing was significantlyaltered by both SP and DAMGO, 21% were sensitive toneither, and 42% were sensitive to one peptide only.

A total of 40 VP neurons were tested for sensitivity toSP; 66% responded to the neuropeptide and all but oneexhibited rate increases. SP-sensitive neurons did not ex-hibit any tendency for clustering either along the rostro-caudal axis or along the ventromedial–dorsolateral axis ofthe VP.

Fig. 5. Illustrations of the effects of SP on spontaneous VP neuronalactivity and short latency excitations evoked by activating the AMG. A.A real time rate histogram illustrating VP neuronal responses elicited byiontophoretically applied SP. The horizontal bars flanked with arrows

Ž .indicate onset and offset of the ejection current. The dashed line stimrepresents the time period during which AMG was stimulated. Thenumbers above the bars indicate SP ejection current magnitude. B.

Ž .Peristimulus time histograms in a different VP neuron 128 epochs eachillustrating the ability of sub-threshold ejection currents of SP to attenuateAMG-evoked short latency excitations. The boxes indicate the onset,duration and peak countsrbin of excitation as defined during control

Ž .condition i.e., before SP iontophoresis; CTRL . SP 20 nA indicates theejection current magnitude used to apply SP during AMG stimulation. Inthis neuron, 65 action potentials occurred in the evoked componentduring CTRL. The expected count during SP was calculated to be 65.5while 51 action potentials actually occurred. Following recovery, 64action potentials were elicited in the short latency response. As reflected

Ž .by the number of action potentials that occurred during CTRL 65 andŽ .the expected count 65.5 , at this ejection current level, SP did not alter

the interstimulus baseline, however, the peptide did suppress the shortlatency excitation in this cell that was mediated by activating the AMG.C. A bar graph summarizing the suppression of the AMG-evoked short

Ž .latency excitations by iontophoretically applied SP. The open bar CTRLrepresents the total number of counts in short latency responses evoked

Ž .by stimulation of the AMG currents ranged from 0.4–1.4 mA . Thestippled bar represents calculated expected number of counts for the shortlatency responses obtained during iontophoresis of sub-threshold currents

Žof SP. 2 to 20 nA; ) p-0.05; responses compared with the expectedw x .number using paired t- test defined in Section 2 for 11 neurons.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206 201

Fig. 6. Illustrations of the effects of DAMGO on spontaneous VPneuronal activity and AMG stimulation-evoked short latency excitatoryresponses of VP neurons. A. A real time rate histogram of neuronalresponses elicited by iontophoretically applied DAMGO. The horizontalbars flanked with arrows indicate onset and offset of the ejection current.The numbers above the bars indicate ejection current. The dashed lineŽ .stim represents the time period of AMG stimulation. B. Peristimulus

Ž .time histograms in a different VP neuron 128 epochs each illustratingthe ability of sub-threshold ejection currents of DAMGO to potentiateAMG-evoked short latency excitation. The boxes indicate the onset,duration and peak countsrbin of excitation as defined during control

Ž .condition i.e., before DAMGO iontophoresis; CTRL . DAMGO 20 nArepresents the current magnitude used to apply the opioid during AMGstimulation. In this neuron, 47 action potentials occurred in the evokedcomponent during CTRL. The expected count was calculated to be 50.1.Sixty action potentials occurred during the excitatory component whileDAMGO was being applied, and 38 actually occurred after recovery. C.A bar graph summarizing the suppression of the short latency responses

Ž .to iontophoretically applied DAMGO. The open bar CTRL representsthe total number of counts in the short latency responses evoked by

Ž .stimulation of the AMG currents ranged from 0.6–1.4 mA . The stippledbar represents the calculated expected number of counts for the shortlatency responses obtained during iontophoresis of sub-threshold currents

Žof DAMGO. 1–25 nA; ) p-0.05; responses compared with the ex-w xpected number using a paired t- test defined in Section 2 for 11

.neurons.

Eleven VP neurons demonstrating short latency excita-tions to AMG stimulation as well as firing rate increases toiontophoretically applied SP were used to test the ability oflocally applied SP to modulate this AMG-evoked excita-tion. The AMG was stimulated during iontophoretic appli-cation of the neuropeptide using ejection currents for SPthat were below the level required to significantly increase

Žspontaneous neuronal activity i.e., less than 20% changefrom the spontaneous firing; see Fig. 5A; for detailed

.description see methods . Fig. 5B illustrates a typicaleffect of SP on the AMG-evoked response. Although theiontophoretically applied tachykinin generally increasedspontaneous VP neuronal activity, at ejection currents thatwere sub-threshold for this effect, SP consistently attenu-ated the AMG-evoked short latency excitations, with an

Ž .average of 37"14% Fig. 5C . For comparison purposes,we also tested the ability of sub-threshold ejection currentlevels of SP to modulate the rate increases elicited by

Ž .iontophoresed exogenous glutamate. In six VP neuronstested, iontophoretically applied glutamate robustly in-creased activity, but in contrast to the ability of SP toattenuate the excitatory effects of stimulating the AMGŽ .which likely reflects an endogenously released glutamate ,SP did not alter the effects of exogenous glutamate on

Ž .these cells ts1.387; ps0.224 . These results would beexpected if SP modulated AMG stimulation-evoked re-lease of glutamate through an action at a pre-synaptic site,but did not alter the effect of glutamate at a post-synapticsite.

Forty VP neurons evoked by stimulation of AMG weretested for sensitivity to iontophoretically applied DAMGO.Two thirds of the tested neurons responded to the opioidand a decrease in neuronal activity was obtained almostexclusively. In contrast to SP, opioid-sensitive neurons

Ž 2 .were more frequently encountered x s4.841; p-0.03in the dorsolateral VP. In 11 VP neurons that exhibitedboth AMG-evoked short latency excitations andDAMGO-induced rate suppressions, the ability of DAMGOto modulate the responses to AMG stimulation was as-sessed. The protocol used to test the modulatory effects ofthe sub-threshold ejection currents of SP was employed

Ž .here for DAMGO see Fig. 6A . Fig. 6B illustrates typicalŽ .effects of these low 1–25 nA ejection currents of

DAMGO on the short latency excitatory responses. Incontrast to SP, the opioid potentiated these AMG-evoked

Ž .excitations by 39"26% Fig. 6C . DAMGO also was ableto potentiate effects of exogenously applied glutamateŽ .tsy3.207; p-0.04 in all five neurons tested, suggest-ing that a post-synaptic site may be involved.

4. Discussion

The present study functionally verified VP glutamateand SP inputs from the AMG and the NA, respectively.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206202

This complimented our previous work characterizing thew xphysiology of the opioid input to the VP from the NA 8 .

The present study also demonstrated the ability oftachykinin and opioid neuropeptides to modulate the VPglutamatergic input from the AMG. SP attenuated, andDAMGO potentiated AMG stimulation-evoked glutamater-gic neurotransmission, which contrasted the ability of SPto enhance, and DAMGO to inhibit spontaneous firing.The interactions between SP and glutamate appeared tooccur at a pre-synaptic site, while post-synaptic sites likelyare involved in the ability of DAMGO to modulate theeffects of glutamate. These results demonstrate a complex-ity of modulatory processes imposed by peptidergic inputsfrom the NA on amygdaloid activation of VP neurons.

Electrical stimulation of the AMG elicited both excita-tory and inhibitory responses from VP neurons. Using

Ž .KYN a glutamatergic antagonist , we demonstrated thatglutamate regulated the short latency excitatory responsesto AMG stimulation. This finding corroborates anatomicaldescriptions of a glutamatergic projection from the AMG

w xto the VP 13 . In some VP neurons, AMG stimulation alsoevoked short latency inhibitory responses. The amygdaloidintercalated cell masses use GABA as a transmitter andprojections from these neurons terminate within the VPw x48 . While the paradigm used to stimulate AMG in the

Žpresent study activates a discrete area with an approxi-. w xmate radius of 500 mm from the electrode tip 32 , the

proximity of the basolateral stimulation site with the inter-w xcalated cell masses 48 allows for the possibility that

sometimes the intercalated nuclei were activated. Thiswould explain the observed short latency inhibitory re-sponses to AMG stimulation.

Long latency responses to AMG stimulation wererecorded in some VP neurons. The AMG projects to the

w xNA 64 in addition to VP. In turn, the VP is a main targetw xregion for NA GABA-ergic neurons 51 . Therefore, it is

possible that some of the inhibitory VP responses to AMGstimulation were mediated via an activation of VPGABA-ergic inputs from the NA. This notion is supported

Ž .by the ability of a local anesthetic procaine infused in theNA to block long latency inhibitions, but not the shortlatency inhibitory or excitatory responses that were evoked

w xin VP neurons by AMG stimulation 61 .Several investigators have evaluated VP responses to

electrical activation of the NA, and all have observed VPw x Ž .excitatory responses 8,28,37 . However, the transmitter s

mediating these responses was not identified. The excita-w xtory neuropeptide SP is found in NA efferents 45 , and

thus we considered that SP may be involved in the rateincreases evoked in VP cells by activating the NA. Aspredicted, the NK antagonist CP96345 attenuated the1

long latency excitatory responses, as well as those evokedby iontophoretically applied SP. Additionally, only thoseneurons that exhibited long latency excitations to the elec-trical stimulation of the NA were sensitive to SP. Thismimicry, and the ability of CP96345 to antagonize the rate

increases to both SP and activation of the AMG, stronglysuggest that the long latency evoked excitation was elicitedby a tachykinin. Because SP-induced effects are mediated

w xthrough the G protein-coupled NK receptor 26 , a re-1

sponse with a delayed onset and a long duration would bew xexpected, even if it is mediated monosynaptically 54 . An

accumbal SP projection to the VP also may be activated byAMG excitatory inputs to the NA, and thus could underliethe long latency excitatory responses observed in some VPcells following AMG activation.

Four VP neurons that demonstrated short latency excita-tions to NA stimulation were tested with, and determinedto not be sensitive to iontophoretically applied SP orCP96345. This suggests that projections containing anexcitatory neurotransmitter other than SP also were acti-vated by NA stimulation. The short onset latency and shortduration of the stimulation-evoked response suggest thatthey were mediated by a transmitter acting through an

w xionotropic receptor such as glutamate 9 . Indeed, ion-tophoresis of glutamate increases VP cell firing with a

w xrapid response onset and offset 41 . The medial prefrontalcortex projection to the VP uses glutamate as a neurotrans-

w xmitter, and this projection passes through the NA 14,55 .Thus, it is likely that these fibers of passage were activatedby stimulating the NA.

w xAs previously reported 8,37 , activation of the NAinhibited the firing of a significant portion of the recordedVP neurons. Opioid peptides and GABA are contained in

w xNA projection neurons 51 . Microiontophoretically ap-plied GABA decreases the spontaneous activity of VPneurons and NA-evoked inhibitions are antagonized by

w xbicuculline 8 suggesting that GABA was involved in theinhibitory responses to accumbal stimulation that wereobserved in the present study. The role of accumbal opi-oid-containing inputs in the regulation of VP neuronalactivity is more complex, and may include rate suppres-

w xsions 34,43 as well as a disinhibition of GABA-mediatedw xeffects 8 .

Several studies have demonstrated that concentrationsof neurotransmitters below the level necessary to changespontaneous neuronal activity can modulate the firing rate

w xeffects of amino acids 4,7,22,58 . We previously demon-strated that such sub-threshold levels are sufficient for SP

w xand DAMGO to act as neuromodulators in the VP 35 .The present study extended this work to reveal that gluta-matergic transmission in the VP can be altered by SP andDAMGO using iontophoretic ejection current levels thatwere below those necessary for the peptides to alter spon-taneous firing. Such findings argue for a neuromodulatoryrole for tachykinins and opioids in excitatory amino acidtransmission within the VP.

The results also revealed that the nature of the modula-tion differs for SP and DAMGO. SP attenuated VP neu-ronal excitations mediated by endogenously released gluta-mate. This effect could occur either if SP altered gluta-mate-mediated effects on the post-synaptic neuron andror

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206 203

if the tachykinin altered the pre-synaptic release of theexcitatory amino acid. Although this experimentalparadigm did not directly determine the pre- vs. post-syn-aptic site of interaction, by comparing the ability of SP toalter AMG-evoked excitations to its ability to alter the

Ž .effects of iontophoretically applied exogenous glutamate,the site of interaction can be inferred. The tachykinin didnot modulate the effects of exogenous glutamate, thusrendering a possibility of an interaction on post-synapticneuron unlikely. Therefore, our data suggest that the abil-ity of the activated AMG to release glutamate was beingaltered by SP. A pre-synaptic locale for NK receptors on1

glutamatergic terminals of AMG inputs to the VP wouldallow SP to directly control the release of glutamate.Lending support to this possibility is the discovery that

w xAMG neurons express NK receptor mRNA 24 , and that1

at least in the striatum, NK receptors have been visual-1w xized on pre-synaptic terminals 21 . SP inhibits the N- and

2q w xQ-type of Ca currents 31 . It’s been demonstrated thatw xthese channels are localized presynaptically 59,60 . N-

and Q-type Ca2q channels are known to be critical forglutamatergic transmission in the hippocampus, most likelyfunctioning to mediate the pre-synaptic release of gluta-

w xmate 12 . Therefore, it is possible that the iontophoreti-cally applied SP used in the present study acted throughpre-synaptically expressed NK receptors to inhibit N- and1

Q-type of Ca2q currents and, thus was able to decreaseAMG stimulation-evoked release of glutamate in the VP.

DAMGO potentiated AMG stimulation-evoked excita-tions and the effects of iontophoretically applied gluta-mate. Because the opioid enhanced responses to bothendogenous and exogenous glutamate, it is likely that theeffects of DAMGO observed during AMG stimulation

Žinvolved an interaction at the post-synaptic site although a.modulation of glutamate release can not be excluded . This

w xconclusion is not without precedent as Chen and Huang 6demonstrated that the activation of post-synaptic m opioidreceptors potentiates excitatory responses of spinal trigemi-nal neurons to activation of post-synaptic NMDA recep-tors.

The direction of modulation of glutamate-induced rateincreases by SP and DAMGO was not predicted by theeffects of the neuropeptides on spontaneous firing. SPincreased spontaneous firing, but AMG-evoked excitationswere attenuated by sub-threshold ejection currents of thetachykinin. On the other hand, DAMGO readily sup-pressed spontaneous firing, but at sub-threshold ejectioncurrents, the opioid potentiated AMG-evoked responses.These results suggest that the two peptides have complexeffects on VP neuronal firing, the mode of which maydepend upon the concentration of the neuropeptide androrthe level of activation of the glutamatergic input fromAMG. The NA is a main source of SP and opioid innerva-

w xtion of the VP 17,62 , thus conferring the NA with abroad spectrum of effects in regulation of VP neuronal

Žactivity although a possibility that some other inputs to

the amygdaloceptive VP neurons may also contain SP.andror opioids can not be ruled out . It appears, therefore,

that the NA can profoundly alter the significance of anexcitatory drive on VP cells by changing the amount ofneuropeptide released from its projections to the VP. Themodulating ability of these neuropeptides also may bealtered by the level of glutamatergic transmission from theAMG. Based on the results from the present study and theliterature on the physiology of SP and m opioids, putativescenarios can be proposed for the transmission modes usedby these peptides in the VP: At a sufficient concentration,SP can increase cell firing by acting on post-synaptic NK1

receptors to suppress an inwardly rectifying Kq currentw x39 . But when activity in excitatory inputs from the AMGis enhanced, neuronal activity in the VP appears to reflectthe consequence of SP acting on pre-synaptic NK recep-1

tors to decrease the presynaptic release of the excitatoryamino acid even at concentrations of SP that are too low todirectly excite a VP neuron. DAMGO can directly sup-press cell firing via a m receptor-mediated increase in Kq

w xconduction 16 . However, it is possible that when post-synaptically located NMDA receptors are activated by anenhancement in glutamatergic transmission from the AMG,sub-threshold concentrations of the m agonist are sufficient

Žto potentiate the NMDA-mediated effects refer to the.above discussion . If true, these scenarios would empower

the AMG with an important influence on the ability of theNA to direct neuronal communication in the VP. Thesescenarios also point to the importance of the relative statesof activity between NA and AMG inputs to the VP and byinference, that shifts in this relationship would be animportant determinant in functional state of VP neurons.

Rarely co-localizing, SP- and enkephalin-containingfibers are thought to represent two distinct inputs to the VPw x51 . We observed that of the neurons sensitive to AMGstimulation and tested with both SP and DAMGO, 42%were sensitive to only one peptide. This suggests that asizable portion of VP neurons may receive inputs from theAMG and only one of these neuropeptidergic systems.However, 37% of the VP neurons tested responded to bothneuropeptides, implying that convergence also occurs be-tween tachykinins and opioids in the VP. If so, then for asub-population of VP cells, the NA may be able to potenti-

Ž . Ž .ate opioids and attenuate SP AMG influences.In this study, only those VP neurons that responded to

AMG stimulation were tested for sensitivity to the neu-ropeptides. In this population, neurons that were sensitiveto SP were not preferentially localized to a particularsub-region of the VP. This finding concurs with our previ-ous studies where randomly encountered VP neurons were

w xexamined 35,45 . However, in contrast to our previousw xstudies with randomly encountered neurons 34,35 , neu-

rons that responded to DAMGO in the present study wereobserved with a higher frequency in the dorsolateral VP.While the reason for the difference in distribution ofDAMGO-sensitive neurons between spontaneously active

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206204

and AMG-evoked cells is not known, an enhanced sensi-tivity of neurons in the dorsolateral VP agrees with thedenser enkephalinergic innervation of this region as com-

w w xxpared to the ventromedial VP for reviews, see 2,66 .These two subregions also differ in their afferent and

w xefferent projection patterns 2,64–66 . The dorsolateral VPreceives inputs primarily from the core and the rostral poleof the NA, and projects to the substantia nigra and subtha-

w xlamic nucleus 64,66 . NA afferents from the accumbalshell terminate in the ventromedial VP and neurons fromthis part of the VP project to the ventral tegmental areaw x66 . These connectivity patterns suggest that the dorsolat-eral VP is primarily involved in regulation of motor behav-ior, while the ventromedial VP may subserve limbic func-

w xtions 18 . It is worth mentioning in this context thatbehavioral studies have demonstrated the importance of

w xVP m opioid receptors in locomotion 3,19,42 . It has beenproposed that the interactions among the AMG, NA and

w xVP critically influence motivation 36 . Thus, accumbalopioid modulation of an AMG excitatory projection to themotor segment of the VP by accumbal opioid inputs maybe of significance for the motoric expression of goal-ori-ented behaviors.

In summary, results from the present study suggest thatthe peptidergic modulation of the effects off excitatoryamino acid within the VP may enable the NA to filterAMG influences and insure that relevant informationreaches the appropriate target. This process likely dependson the activation level of both the accumbal peptidergicinputs and the excitatory amino acid inputs from the AMG.Disturbances of this modulation may have particular im-portance in pathophysiology of schizophrenia and Parkin-son’s disease, which are associated with disregulation inglutamatergic transmission within the ventral striatopalli-

w xdal system 5 .

Acknowledgements

The authors extend their gratitude to Drs. P.I. Johnsonand M. Margeta-Mitrovic for their valuable comments onthis manuscript. We also are grateful to Pfizer for thegenerous gift of CP96345. Work was supported byUSPHSG DA05255 to TCN.

References

w x1 J.P. Aggleton, The contribution of the amygdala to normal andŽ .abnormal emotional states, Trends Neurosci. 16 1993 323–328.

w x2 G.F. Alheid, L. Heimer, New perspectives in basal forebrain organi-zation of special relevance for neuropsychiatric disorders: The stri-atopallidal, amygdaloid, and corticopetal components of the substan-

Ž .tia innominata, Neuroscience 27 1988 1–39.

w x3 M.C. Austin, P.W. Kalivas, Enkephalinergic and GABAergic modu-lation of motor activity in the ventral pallidum, J. Pharmacol. Exp.

Ž .Ther. 252 1990 1370–1377.w x4 D.A. Bergstrom, J.R. Walters, Dopamine attenuates the effects of

GABA on single unit activity in the globus pallidus, Brain Res. 310Ž .1984 23–43.

w x5 M. Carlsson, A. Carlsson, Interactions between glutamatergic andmonoaminergic systems within the basal ganglia—implications for

Ž .schizophrenia and Parkinson’s disease, Trends Neurosci. 13 1990272–276.

w x6 L. Chen, L.M. Huang, Sustained potentiation of NMDA receptor-mediated glutamate responses through activation of protein kinase C

Ž .by a mu opioid, Neuron 7 1991 319–326.w x7 L. Chiodo, T.W. Berger, Interactions between dopamine and

amino-acid induced excitation and inhibition in the striatum, BrainŽ .Res. 375 1986 198–203.

w x8 J.J. Chrobak, T.C. Napier, Opioid and GABA modulation of accum-Ž .bens-evoked ventral pallidal activity, J. Neural Trans. 93 1993

123–143.w x9 G.L. Collingridge, A.J. Leseter, Excitatory amino acid receptors in

Ž .the vertebrate central nervous systems, Pharmacol. Rev. 40 1989143–210.

w x10 A.D. Corbett, S.J. Paterson, H.W. Kosterlitz, Selectivity of ligandsŽ .for opioid receptors, in: A. Herz Ed. , Opioids I, Springer-Verlag,

Berlin, 1993, pp. 645–679.w x11 A.R. Crossman, R.J. Walker, G.N. Woodruff, Problems associated

with iontophoretic studies in the caudate nucleus and substantiaŽ .nigra, Neuropharmacology 13 1974 547–552.

w x12 E.M. Elliott, A.T. Malouf, W.A. Catterall, Role of calcium channelsubtypes in calcium transients in hippocampal CA3 neurons, J.

Ž .Neurosci. 15 1995 6433–6444.w x13 T.A. Fuller, F.T. Russchen, J.L. Price, Sources of presumptive

glutamergic–aspartergic afferents to the rat ventral striatopallidalŽ .region, J. Comp. Neurol. 258 1987 317–338.

w x14 R.P.A. Gaykema, R. Van Weeghel, L.B. Hersh, P.G.M. Luiten,Prefrontal cortical projections to the cholinergic neurons in the basal

Ž .forebrain, J. Comp. Neurol. 303 1991 563–583.w x15 H.J. Groenewegen, F.T. Russchen, Organization of the efferent

projections of the nucleus accumbens to pallidal, hypothalamic, andmesencephalic structures: A tracing and immunohistochemical study

Ž .in the cat, J. Comp. Neurol. 223 1984 347–367.w x16 T.J. Grudt, J.T. Williams, Opioid receptors and the regulation of ion

Ž .conductances, Rev. Neurosci. 6 1995 279–286.w x17 S.N. Haber, W.J.H. Nauta, Ramification of the globus pallidus in the

rat as indicated by patterns of immunohistochemistry, NeuroscienceŽ .9 1983 245–260.

w x18 L. Heimer, G.F. Alheid, Piecing together the puzzle of basal fore-Ž .brain anatomy, in: T.C. Napier, P.W. Kalivas, I. Hanin Eds. , The

Basal Forebrain: Anatomy to Function, Adv. Exp. Med. Biol., Vol.295, Plenum Press, New York, 1991, pp. 1–42.

w x19 D.C. Hoffman, T.E.G. West, R.A. Wise, Ventral pallidal microinjec-tions of receptor-selective opioid agonists produce differential ef-fects on circling and locomotor activity in rats, Brain Res. 550Ž .1991 205–212.

w x20 E. Irle, H.J. Markowitsch, Afferent connections of the substantiainnominatarbasal nucleus of Meynert in carnivores and primates, J.

Ž .Hirnforsch. 27 1986 343–367.w x21 R.L. Jakab, P. Goldman-Rakic, Pre-synaptic and post-synaptic sub-

cellular localization of substance P receptor immunoreactivity in theŽ .neostriatum of the rat and rhesus monkey Macaca mulatta , J.

Ž .Comp. Neurol. 369 1996 125–136.w x22 P.I. Johnson, T.C. Napier, Morphine modulation of GABA- and

glutamate-induced changes of ventral pallidal neuronal activity,Ž .Neuroscience 77 1997 187–197.

w x23 A.E. Kelley, V.B. Domesick, W.J.H. Nauta, The amygdalostriatalprojection in the rat—an anatomical study by anterograde and

Ž .retrograde tracing methods, Neuroscience 7 1982 615–630.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206 205

w x Ž .24 H. Kiyama, H. Maeno, M. Tohyama, Substance P receptor NK1 inthe central nervous system: Possible functions from a morphological

Ž .aspect, Regul. Pept. 46 1993 114–123.w x25 G.F. Koob, Drugs of abuse: Anatomy, pharmacology and function of

Ž .reward pathways, Trends Pharmacol. Sci. 13 1992 177–184.w x26 J.E. Krause, B.S. Sachais, P. Blount, Tachykinin receptors, in: S.J.

Ž .Peroutka Ed. , Handbook of Receptors and Channels, CRC Press,Boca Raton, 1993, p. 227.

w x27 R.A. Lahti, M.M. Mickelson, K.S. Jodelis, J.M. McCall, Compara-tive neuroanatomical distribution of the kappa and mu opioid

Ž .receptors in guinea pig brain sections, Eur. J. Pharmacol. 166 1989563–566.

w x28 A. Lavin, A.A. Grace, Physiological properties of rat ventral pal-lidum neurons recorded intracellularly in vivo, J. Neurophys. 75Ž .1996 1432–1433.

w x29 J. Lipiski, Antidromic activation of neurons as an analytic tool in theŽ .study of the central nervous system, J. Neurosci. Meth. 4 1991

1–32.w x30 A. Mansour, C.A. Fox, R.C. Thompson, H. Akil, S.J. Watson,

m-Opioid receptor mRNA expression in the rat CNS: Comparison toŽ .m-receptor binding, Brain Res. 643 1994 245–265.

w x31 M. Margeta-Mitrovic, J.J. Grigg, K. Koyano, Y. Nakajima, S.Nakajima, Neurotensin and substance P inhibit low- and high-volt-age-activated Ca2q channels in cultured newborn rat nucleus basalis

Ž .neurons, J. Neurophys. 78 1997 1341–1352.w x32 R.J. Maslowski-Cobuzzi, T.C. Napier, Activation of dopaminergic

neurons modulates ventral pallidal responses evoked by amygdalaŽ .stimulation, Neuroscience 62 1994 1103–1120.

w x33 A.J. McDonald, J.L. Culberson, Efferent projections of the basolat-eral amygdala in the opossum, Didelphis Õirginiana, Brain Res.

Ž .Bull. 17 1986 335–350.w x34 I. Mitrovic, T.C. Napier, Electrophysiological demonstration of mu,

delta and kappa opioid receptors in the ventral pallidum, J. Pharma-Ž .col. Exp. Ther. 272 1995 1260–1270.

w x35 I. Mitrovic, T.C. Napier, Interactions between the mu opioid agonistDAMGO and substance P in the regulation of the ventral pallidum,

Ž .Synapse 23 1996 142–151.w x36 G.J. Mogenson, C.R. Yang, The contribution of basal forebrain to

limbic-motor integration and mediation of motivation to action, in:Ž .T.C. Napier, P.W. Kalivas, I. Hanin Eds. , The Basal Forebrain,

Anatomy to Function, Adv. Exp. Med. Biol., Vol. 295, PlenumPress, New York, 1991, pp. 267–290.

w x37 G.J. Mogenson, L.W. Swanson, M. Wu, Neural projections fromnucleus accumbens to globus pallidus, substantia innominata, andlateral preoptic-lateral hypothalamic area: an anatomical and electro-

Ž .physiological investigation in the rat, J. Neurosci. 3 1983 189–202.w x38 A.S. Moskowitz, R.R. Goodman, Light microscopic autoradio-

graphic localization of mu and delta opioid binding sites in theŽ .mouse central nervous system, J. Neurosci. 4 1984 1331–1342.

w x39 Y. Nakajima, K. Koyano, B.M. Velimirovic, J.J. Grigg, S. Naka-jima, Signal transduction mechanisms of substance P effects on ion

Ž .channels, Regul. Pept. 46 1993 110–113.w x40 T.C. Napier, J.H. Pirch, H.K. Strahlendorf, Naloxone antagonizes

striatally-induced suppression of globus pallidus unit activity, Neu-Ž .roscience 9 1983 53–59.

w x41 T.C. Napier, P.E. Simson, B.S. Givens, Dopamine electrophysiologyof ventral pallidalrsubstantia innominata neurons: Comparison with

Ž .the dorsal globus pallidus, J. Pharmacol. Exp. Ther. 258 1991249–262.

w x42 T.C. Napier, Dopamine receptors in the ventral pallidum regulatecircling induced by opioids injected into the ventral pallidum, Neu-

Ž .ropharmacology 31 1992 1127–1136.w x43 T.C. Napier, J.J. Chrobak, J. Yew, Systemic and microiontophoretic

administration of morphine differentially effect ventralŽ .pallidumrsubstantia innominata neuronal activity, Synapse 12 1992

214–219.w x44 T.C. Napier, Transmitter actions and interactions on pallidal neu-

Ž .ronal functions, in: P.W. Kalivas, C.D. Barnes Eds. , Limbic MotorCircuits and Neuropsychiatry, CRC Press, Boca Raton, 1993, pp.125–153.

w x45 T.C. Napier, I. Mitrovic, L. Churchill, M.A. Klitenick, X.Y. Lu,P.W. Kalivas, Substance P in the ventral pallidum: Projection fromthe ventral striatum, and electrophysiological and behavioral conse-

Ž .quences of pallidal substance P, Neuroscience 69 1995 59–70.w x46 S. Oleskevich, J.D. Clements, J.T. Williams, Opioid-glutamate inter-

Ž .action in rat locus coeruleus neurons, J. Neurophysiol. 70 1993931–937.

w x47 D. Olton, A. Markowska, M.L. Voytko, B. Givens, L. Gorman, G.Wenk, Basal forebrain cholinergic system: a functional analysis, in:

Ž .T.C. Napier, P.W. Kalivas, I. Hanin Eds. , The Basal Forebrain:Anatomy to Function, Adv. Exp. Med. Biol., Plenum Press, NewYork, 1991, pp. 353–372.

w x48 D. Pare, Y. Smith, GABAergic projection from the intercalated cell´masses of the amygdala to the basal forebrain in cats, J. Comp.

Ž .Neurol. 344 1994 33–49.w x49 G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Coordinates,

Academic Press, New York, 1986.w x50 C. Pilapil, S. Welner, J. Magnan, S. Gauthier, R. Quirion, Autora-

diographic distribution of multiple classes of opioid receptor bindingŽ .sites in human forebrain, Brain Res. Bul. 19 1987 611–615.

w x51 A. Reiner, K.D. Anderson, The patterns of neurotransmitter andneuropeptide co-occurrence among striatal projection neurons: con-

Ž .clusions based on recent findings, Brain Res. Rev. 15 1990 251–265.

w x52 D. Regoli, A. Boudon, J.-L. Fauchere, Receptors and antagonists for´Ž .substance P and related peptides, Pharmacol. Rev. 46 1994 551–

599.w x53 K.I. Rusin, P.D. Ryu, M. Randic, Modulation of excitatory amino

acid responses in rat dorsal horn neurons by tachykinins, J. Neuro-Ž .physiol. 68 1992 256–286.

w x54 J.H. Schwartz, E.R. Kandel, Synaptic transmission mediated bysecond messengers, in: E.R. Kandel, J.H. Schwartz, T.M. JessellŽ .Eds. , Principles of Neural Science, Elsevier Science Publishing,New York, 1991, pp. 173–193.

w x55 S.R. Sesack, A.Y. Deutch, R.H. Roth, B.S. Bunney, Topographicalorganization of the efferent projections of the medial prefrontalcortex in the rat: An anterograde tract-tracing study with Phaseolus

Ž .Õulgaris leucoagglutinin, J. Comp. Neurol. 290 1989 213–242.w x56 R.M. Snider, J.W. Constantine, J.A. Lowe, K.P. Longo, W.S. Lebel,

H.A. Woody, S.E. Drozda, M.C. Desai, F.J. Vinick, R.W. Spencer,Ž .H.-J. Hess, A potent nonpeptide antagonist of the substance P NK1

Ž .receptor, Science 251 1991 435–437.w x57 C.T. Tsai, G.J. Mogenson, M. Wu, C.R. Yang, A comparison of the

effects of electrical stimulation of the amygdala and hippocampus onsubpallidal output neurons to the pedunculopontine nucleus, Brain

Ž .Res. 494 1989 22–29.w x58 B.L. Waszczak, J.R. Walters, Dopamine modulation of the effects of

g-aminobutyric acid on substantia nigra pars reticulata neurons,Ž .Science 220 1983 218–221.

w x59 R.E. Westenbroek, J.W. Hell, C. Warner, S.J. Dubel, T.P. Snutch,W.A. Catterall, Biochemical properties and subcellular distribution

Ž .of an N-type calcium channel a subunit, Neuron 9 1992 1099–1

1115.w x60 R.E. Westenbroek, T. Sakurai, E.M. Elliott, J.W. Hell, T.V.B. Starr,

T.P. Snutch, W.A. Catterall, Immunochemical identification andsubcellular distribution of the a subunits of brain calcium chan-1

Ž .nels, J. Neurosci. 15 1995 6403–6418.w x61 C.Y. Yim, G.J. Mogenson, Response of ventral pallidal neurons to

amygdala stimulation and its modulation by dopamine projections toŽ .nucleus accumbens, J. Neurophysiol. 50 1983 148–161.

w x62 L. Zaborszky, G.F. Alheid, L. Heimer, Mapping of transmitter-´specific connections: Simultaneous demonstration of anterogradedegeneration and changes in the immunostaining pattern induced by

Ž .lesions, J. Neurosci. Meth. 14 1985 225–266.

( )I. MitroÕic, T.C. NapierrBrain Research 792 1998 193–206206

w x63 L. Zaborszky, C.S. Leranth, L. Heimer, Ultrastructural evidence ofamygdalofugal axons terminating on cholinergic cells of the rostral

Ž .forebrain, Neurosci. Lett. 52 1984 219–225.w x64 D.S. Zahm, The ventral striatopallidal parts of the basal ganglia in

the rat: II. Compartmentalization of ventral pallidal efferents, Neuro-Ž .science 30 1989 33–50.

w x65 D.S. Zahm, J.S. Brog, On the significance of subterritories in theŽ .‘accumbens’ part of the rat ventral striatum, Neuroscience 50 1992

751–767.w x66 D.S. Zahm, L. Heimer, Two transpallidal pathways originating in

Ž .the rat nucleus accumbens, J. Comp. Neurol. 302 1990 437–446.